(To the tune of the “All in the Family” theme song): “‘Disco Duck’ and Fleetwood Mac, coming out of my eight-track, / Michael Jackson still was black. / Those were the daaaaaaays!”
– “The Simpsons”
Those of you who peruse this space each week have no doubt noticed my recurrent use of such phrases as “back in the days when….” or “way back when…” or “in the beginning…”
This is because I constantly find myself revisiting the past in order to make sense of the future, which in the Internet universe finds itself very quickly becoming the present.
So let me begin.
For the last few years, most online advertising focused on one of two objectives:
- Getting the name out and driving some traffic regardless of sales or registrations – making a massive splash on the web. The “ubiquity tactic,” I call it. This is how Saturn started when Hal Riney & Partners first took them on the web.
- Direct response: moving as many widgets as possible as efficiently as possible.
Of the two, the direct response objective is the more common. Almost all online advertisers want some result. They want emails, registrations, demo downloads, or products purchased.
But regardless of whether you are interested in the result of simple traffic or complex transactions, there is one thing that remains constant: testing.
I’ve written in this space before about the importance of testing and how both basic and nuanced decisions would be based on the reads of data gathered from these tests. In all cases, it comes down to either efficiencies or volumetrics that determine which properties you continue to advertise on and which you don’t. And usually, efficiencies and volume go hand-in-hand (though not always… this is not an axiom).
And so, for the first few years, states of perennial testing were maintained. And given the marketplace at the time, where there was lots of inventory everywhere and sites and publishers were all fighting over the same nickel, an advertiser could sustain phases of long-term testing without ever having to commit to anything beyond the month of the buy.
However, once you’ve tested and tested again, you really have to do something meaningful with the results. There will finally come a time to decide on a long-term commitment.
Also, it is no longer just a buyer’s market. As more and more advertisers enter the marketplace, and there is more and more competition for inventory, sites and their publishers are no longer as amenable to a constant state of test buys. There are now advertisers out there who are willing to buy up inventory for longer schedules without testing just to be sure they can hold on to it if it looks like it’s going to pay out.
Long-term buying strategies have to become more and more a part of the planner and buyer’s everyday consideration. With so much competition and slower growth of inventory than before, rates are going up, and availability is going down. You have to be willing to look at long-term commitments as a rule rather than an exception.
But when is the right time to make this kind of commitment, and what should you look for?
Well, I’ve talked about this in a number of different ways when discussing insertion orders and umbrella agreements and portal deals, but here’s what you should do and when:
- The heat of the advertiser’s category and the threat of scant availability should not preclude a testing phase. You should still do at least a month of testing on as many properties as you can in as many categories as is appropriate. Instinct and fear, though necessary conditions for making long-terms buys, are hardly sufficient conditions for them.
- Analyze the data as soon as there is enough to be statistically valid for the purposes of prediction. That could be one month’s worth or one week’s. Though I prefer to test over at least two months, as some sites experience burnout and tank in the second month, sometimes the competitive set prevents that kind of time and diligence. That means you should probably check on your campaign regularly once a day to see how performance is doing. Once you feel you’ve got enough data on which to base some projections, do it. Then pick your winners and start looking at long-term buy options.
- Once you’ve got your read and made your selections, start working closely with the property on rates and terms and conditions. Because you are willing to commit to a longer schedule, sites should be willing to give you certain rate breaks. You are making their lives easier (think of the phone calls the reps won’t have to make) and you represent the commitment of future cashflow. It should be worth it to them to cut you a deal.
- Look to establish “out” clauses. Not every deal works out for the long term. Make sure you get something in writing that says if performance slips, you can cancel the contract with a certain amount of lead time. Depending on the length of the term, seven to 30 days is standard.
- If you are unsure about a site, then be sure you’ve developed a good relationship with your rep there. He or she can keep you abreast of developments with the property and even who might be sniffing around for inventory you or your client might be interested in. And besides, a good relationship with a rep never hurt anyone.
The industry is red-hot, and there doesn’t seem to be any end in sight for now. New business models are being born every day, and new companies based on them emerge with money to spend. The competition is only going to get tougher. Be prepared to get tough with it.